Pave

Paving the way to better state management.

Why?

Pave is a data layer inspired by Relay, GraphQL, Falcor and Om (next). Pave attempts to take the best pieces of each (subjectively) and expose a simple API that makes managing app state much easier.

Goals

• Performance The core of any data layer will have many hot paths and should run efficiently.

• Flexible Schema There is no requirement that your data is strictly typed, but you can add constraints on inputs and outputs as you see fit.

• POJO All data is represented as JSON-friendly Plain Ol' JavaScript Objects so there is no need to worry about how to serialize X and how to deserialize Y. There are no classes in Pave.

• Multiple Remotes Create a client side schema to use Pave with an existing REST API, implement a Pave schema on the server or mix and match. Allowing multiple remotes both on the client and server makes integrating Pave into an existing project manageable.

• Immutable The cache accepts updates without mutating previous states. This makes history tracking trivial and prohibits unexpected mutations.

• Small pave

  graphql

  falcor

  react-relay

  @apollo/client

Install

npm install pave